Here at UC Berkeley, our central event calendar allows event contributors to type or paste in free form text, which comes from word processing programs, web pages, and the like.
Frequently that text includes non-US-ASCII characters. Typical examples are curly single or double quote marks, em- and en-dashes, characters from non-English languages (e.g. the Spanish 'ntilde' character), and characters from OS-specific character sets.
For those of you who may have encountered this in your own events calendars, have you evolved any best practices for a) filtering or converting such characters, and/or b) encoding them in representations of your events, such as iCalendar or XML (e.g. Atom) feeds?
Stated another, perhaps more geeky, way:
For text-valued fields of your calendar events, such as those corresponding to such iCalendar properties as SUMMARY, DESCRIPTION, and LOCATION, what guidelines or advice would you suggest to help ensure that text is interoperable among various calendaring systems?
In other words, if a calendar event contributor includes an en-dash or a tilde in her description, are there best practices to ensure that character can be viewed as it was originally entered in your customers' Atom or RSS feeds, or in your web-based or desktop (e.g. Outlook, Oracle Calendar) calendaring clients?